Electronic systems and components, particularly those that operate in radio-frequency (RF) ranges, are sensitive to the physical size of the constituent components and interconnects. Thus, changes in the geometry or layout of a transmission line, capacitor, antenna, tuning stub, filter, or other components will affect the performance and/or operating frequency of the RF system. These geometrical features are typically viewed as permanent characteristics of a system once the design, fabrication, and assembly is completed. In order to optimize the system after construction, engineers often tune components by physically removing metal from components (a typical approach for tuning planar antennas), adding bond wires, turning tuning screws, or changing lengths of other adjustable interconnections. These methods are not only time-intensive, but require manual implementation each time a change is desired. If the RF electronic system is installed in an environment that influences its performance, then additional tuning after installation is often required. Conventional methods for physically adjusting the topology or layout of a system are not dynamic and do not enable dynamic adjustments to the system.
In order to allow for dynamic tuning (as opposed to geometric reconfiguration), RF engineers often use tunable capacitors and/or electrical switches. Such devices allow engineers to adjust the system for environmental changes and also give a new capability to provide real-time steering, tuning, band-switching, and other changes to the RF system. However, these lumped-element individual components introduce performance drawbacks including electrical loss, and also require power to hold a given configuration.
In the specific case of RF antennas, due to the limitations of conventional technologies, today's personal and miniature communications systems generally over-specify the physical size, spectral bandwidth, and/or aerial coverage of the antenna. A great deal of effort in the microwave community is being dedicated to antenna miniaturization, but generally only towards minimizing the antenna footprint. Typically, the thickness of the antenna substrate remains unchanged, which is problematic for ultra-miniature applications. The research community has not in general taken on the challenge of reducing this dimension since doing so tends to reduce the antenna bandwidth.
In general, antenna bandwidth is over-specified so that even if the antenna is detuned, it will still capture the desired signal band. The signal bandwidth for the commercial CA-code GPS signal is only about 2 MHz wide but typical GPS patch antennas have a bandwidth of 20 MHz or more to accommodate temperature variations and proximity effects. The bandwidth of a patch antenna is roughly proportional to the substrate thickness, so that if the antenna can be kept on frequency with an active tuning system, the bandwidth and substrate thickness may be reduced by an order of magnitude. Standard approaches to antenna tuning (like those described above), however, generally degrade the antenna's performance. This is because antennas that are tuned via adjustable loading must typically be designed for operation confined to some portion of the tunable band, thereby degrading efficiency. The ability to tune the actual antenna geometry is not typically pursued.
Miniature antennas are, generally, also not adaptable to spatial variations in the external signal. The spatial signal profile and its polarization can vary dramatically due to “multipath” propagation and other environmental effects. Directionally specific, steerable antennas have long been used in complex systems where power and space are available for both the antenna and associated components to operate it. Both analog and digital approaches (e.g., beamforming) may be used, even in strapped-down cellular base stations. However, steerable antennas have generally not been leveraged into miniature communications systems due to the complexity of mechanical and electrical support.
As a result, communications systems generally transmit orders of magnitude more power and require more spectrum usage than they would if it were possible to stay within the signal bandwidth and transmit to the precise location needed.
A common method of adjusting narrow-bandwidth patch antennas is to add solid tuning “fingers” to the edges of the patch. The fingers are usually trimmed by hand with a knife. The size and number of fingers may be selected to allow for very fine control of the patch frequency and input impedance even with relatively coarse adjustments to the length of the fingers. Again, however, such tuning fingers are themselves not dynamically adjustable, and provision and/or manual trimming of such fingers a multitude of times is unwieldy and impractical.
The use of varactor diodes to tune a microstrip patch antenna has also been explored. Numerous researchers have expanded on this approach with multiple-diode configurations, and other antenna geometries. However, although well-suited to receive applications, varactor diodes are highly non-linear and can generate significant unwanted harmonics even at moderate transmit power levels. More recently, MEMS varactors have also been applied in antenna tuning applications. In order to achieve larger frequency shifts for band-switching applications, changes in polarization or antenna pattern, PIN diodes, FETs, and MEMS switches have been used. However, these and other techniques typically produce discrete steps in performance, not continuous, analog tuning, and typically require applied power to maintain a specific configuration.
In many RF and power systems, tunable circuit elements are valuable for optimizing system performance in a dynamic environment (such as matching in response to environmental pulling). Tunable circuit elements may also be beneficial to systems having dynamically changing requirements (such as systems having changing power levels) or to reconfigurable systems (such as those that allow communication channels to be changed). Tunable circuit elements may even facilitate optimizing performance after installation. Even though RF and power systems typically include both inductors and capacitors, at present tunable inductors are not generally available. As such, many existing systems rely almost entirely on tunable capacitors and/or switches, or tunable guided-wave structures such as tunable stubs implemented with switching elements. This generally limits the topologies available for tuning a system. In addition, by only being able to tune a capacitance and not an inductance, the ability to optimize a system for all performance parameters simultaneously (such as frequency of operation and impedance) is often limited.